Blast radius: what a change could break beyond its diff, proven by running code; 'blast radius of X', "이거 바꾸면 뭐가 깨져", "영향 범위 봐줘", a small diff you don't trust. Interrogate: "interrogate", "adversarial review", "challenge this", "find blind spots" on a design, plan or decision. A PR headed to main is deliver-ticket's review.
Collected from public GitHub. Every version is saved exactly as it was and never changes. Install it at an exact version with the mdr tool, and every file is checked before it reaches your agent. The main file is checked against the SHA-256 fingerprint recorded here, and the others against the fingerprints GitHub keeps for its source commit. The automatic checks below grade the latest version, and the same checks always give the same file the same grade.
What the file says
# Challenge
Attack a change or a plan before it ships. Pick the mode by what is being challenged, then read its file and follow it:
- **blast-radius**: a change, often a small diff you don't trust. What it breaks beyond the diff, and the one fact it is safe because of, proven by running real code. [references/blast-radius.md](references/blast-radius.md).
- **interrogate**: a design, plan, decision or diff put to independent reviewers. One verdict sorted into Act on, Consider, Noted and Dismissed. [references/interrogate.md](references/interrogate.md).
A design objection does not finish a blast-radius run: the safety fact is proven by code that ran, or marked unproven. The verdict buckets, which `deliver-ticket` also sorts its review findings by, are in [references/lead-judgment.md](references/lead-judgment.md). The independent review of a PR headed to main is `deliver-ticket`'s, not this skill's.
The line at the top of this page pins the label git:20261107.1b5aba1, which is what most people want. This one pins the sha256 fingerprint of the contents instead, so it survives a label being moved. Either way the exact version is written to mdr.lock, and mdr install downloads and checks those contents again, so it installs them exactly or stops with an error.
GET https://markdownregistry.com/api/v1/artifacts/art_2x2u353qyliq6qw3
GET https://markdownregistry.com/api/v1/resolve?ref=jeongjaesoon/agent-skills/challenge
GET https://markdownregistry.com/api/v1/blob/014cecb62944aaa976a147adc46b25c895c0be4e2bc7cddc85f8376fbdaaac7e
Your AI assistant does the legwork. You hear about the deals worth your word. Hand yours the standing instructions at modelranch.com and it joins the network that reads files like this one.
jeongjaesoon/agent-skills · skills/architect/SKILL.md · Sketch types, signatures and module structure before code, then stay in the loop as it fills in. Use for /architect…
jeongjaesoon/agent-skills · skills/brief/SKILL.md · Status report: "현황 보고해줘", "상태 보고해줘", "브리핑해줘", "진행 상황 알려줘", "진행상황 표로", `/brief`, a turn ending in a report on several…
jeongjaesoon/agent-skills · skills/deliver-ticket/SKILL.md · Use when working a ticket from first edit to done: a PR opened, updated or merged ("PR 올려줘", "머지해줘", "머지까지 진행해줘", "머지되면…
jeongjaesoon/agent-skills · skills/dispatch-card/SKILL.md · Use to start a new Orca card: work for another repo or a side branch while this session goes on — "별도 세션으로 진행해줘", "~…
jeongjaesoon/agent-skills · skills/end-session/SKILL.md · Use to end, close or archive this session, card or worktree — "종료해줘", "세션 종료해줘", "현재 세션 정리해줘", "아카이브해줘", "머지하고 종료하자"…
jeongjaesoon/agent-skills · skills/explain-code/SKILL.md · Before changing unread code: '이 구조 설명해줘', 'X는 어떻게 동작해', '이건 어디에 둬야 해', '왜 이렇게 됐어', '이 결정 배경이 뭐야', 'teach me this', '이거…
jeongjaesoon/agent-skills · skills/fan-out/SKILL.md · Fan out parallel workers. Sweep splits one question, returns one report: /swarm, "병렬로 훑어줘", "나눠서 동시에 확인해줘". Compete…
jeongjaesoon/agent-skills · skills/figure-it-out/SKILL.md · Design an auditable playbook before code for a large migration or multi-part change, or work reviewed after the human…
jeongjaesoon/agent-skills · skills/orchestrate/SKILL.md · Use when this session is the top-level orchestrator ("오케스트레이터로", "모든 세션 관리해줘", "이 세션도 편입해줘") or drives a program to…
jeongjaesoon/agent-skills · skills/principles/SKILL.md · Use when a design, refactor, verification or delegation call needs a named engineering principle: how big a diff, where…
alirezarezvani/claude-skills · c-level-advisor/executive-mentor/skills/challenge/SKILL.md · Use before committing serious resources to a plan, or when feedback has been suspiciously positive. Runs a pre-mortem…
pedrohcgs/claude-code-my-workflow · .claude/skills/challenge/SKILL.md · Stress-test a finding against the choices you did not make. Enumerates the discrete forks a competent analyst could…
oliver-kriska/claude-elixir-phoenix · plugins/elixir-phoenix/skills/challenge/SKILL.md · Challenge mode reviews - rigorous questioning before approving changes. Use when you want thorough scrutiny of Ecto…
me2resh/apexyard · .claude/skills/challenge/SKILL.md · Invoke The Contrarian (Naqid) to steelman-then-challenge an idea, feature, decision, or plan — surfaces assumptions…
athola/claude-night-market · plugins/gauntlet/skills/challenge/SKILL.md · Presents adaptive codebase challenge questions with multiple-choice and trace exercises. Use when testing contributor…
agentii-ai/agentii-investment-intelligence · plugins/vertical-plugins/scenarios/skills/agentii/challenge/SKILL.md · Adversarial verification of research theses — cross-run/cross-thesis contradiction via the entity index, pre-mortem…
